| Abstract | Corporations’ decision-making would be distorted by high corporation taxes, furthermore, affecting the healthy development of corporations, especially for those private corporations which have becoming the new engine of economic growth. Meanwhile, identifying and clarifying the distribution of tax burden among productivity factors is rather meaningful for government to properly make tax reforms and soundly transform governmental functions. Based on a natural experiment of value-added tax reform in Dongbei district, this paper applies the benchmarking approach—difference-in-differences (DiD), to examine the tax burden distribution between corporation and employees. After arguing the rationality of DiD model and “the parallel assumption” in details, we use two different estimation techniques in dynamic panel specification—Diff-GMM and Sys-GMM developed by Arellano & Bond (1991) and Blundell & Bond, respectively. We find a significant and positive wage effect of VAT reform in Dongbei district which suggests that wages in post-reform period were on average 7% higher, for Liaoning province only, the number increased to 8.9%, all these facts suggesting tax-cut profit would share between corporation and employees, i.e., labor certainly will bear a substantial burden of the corporation tax. Then we translate this relationship into money value, that is $1 increase in tax will associate with a decline of 0.44$ in wage.We also use different control groups to make our conclusion more robust. |